Post on 13-Jul-2020
transcript
Material-Process-Performance Relationships in PEM Catalyst Inks and Coated LayersP.I.: Michael UlshPresenter: Scott MaugerNational Renewable Energy LaboratoryMay 21, 2020
DOE Hydrogen and Fuel Cells Program 2020 Annual Merit Review and Peer Evaluation Meeting
This presentation does not contain any proprietary, confidential, or otherwise restricted information.
TA008
NREL | 2
Overview
• Project start date: 10/1/16• FY19 DOE funding: $300,000• FY20 planned DOE funding:
$ 300,000– NREL: $270,000– CSM: $30,000
Timeline and Budget
Barriers
• Colorado School of Mines– Svitlana Pylypenko
Funded Partners
Barrier Target
A. Lack of high-volume MEA processes
$20/kW (2020) at 500,000 stacks/yr
H. Low levels of quality control
NREL | 3
Relevance: Project Addresses MYRD&D Plan Milestones
Task 5: Quality Control and Modeling and Simulation
5.5 Develop correlations between manufacturing parameters and manufacturing variability, and performance and durability of MEAs. (4Q, 2018)
Task 1: Membrane Electrode Assemblies
1.2 Develop processes for direct coating of electrodes on membranes or gas diffusion media. (4Q, 2017)
1.3 Develop continuous MEA manufacturing processes that increase throughput and efficiency and decrease complexity and waste. (4Q, 2017)
• Roll-to-roll (R2R) is expected to be the lowest cost/highest throughput method for MEA mass production, but it has yet to be proven that these methods can produce components meeting performance requirements
• R2R coating techniques require different ink formulation and have different physics than lab-scale processes
• Many researchers/producers do not have access to the infrastructure to understand how the conditions and processes of R2R will impact their materials
• Results directly relevant to researchers and producers• Provides a knowledge and equipment platform for national lab and
university materials development.
NREL | 4
Relevance: Project Success Has Led to Additional DOE Projects
• AMO Roll-to-Roll Consortium (TA007)– Lead for Fuel Cell Core Lab Project – Lead for CRADA with Nel (joint funding from AMO and FCTO)
• HydroGen (PD148)– LTE/Hybrid Supernode (P148a)– Nel Hydrogen project (P155)
• ElectroCat (FC160)• HyET H2@Scale CRADA (H2006)
– “Membrane Electrode Assembly Manufacturing Automation Technology for the Electrochemical Compression of Hydrogen”
• 3M FY19 FOA Award (TA026)– “Low-cost, High Performance Catalyst Coated Membranes for PEM
Water Electrolyzers” • Nel Hydrogen FY19 FOA Award (TA036)
– “Advanced Electrode Manufacturing to Enable Low Cost PEM Electrolysis”
• Lynntech TSA – completed in 2019• Giner TSA – upcoming
NREL | 5
Approach: Study Transition from Lab-Scale to Scalable Electrode Production
Lab Scale – Ultrasonic Spray Large Scale – Roll-to-Roll (R2R)
Conditions• Speed – cm2/min• Dilute ink (~0.6 wt% solids)• Ultrasonic mixing• Sequential build up of layers• Heated substrate• Vacuum substrate
Conditions• Speed – 10s m2/min• Concentrated ink (~4.5-15 wt% solids)• Shear mixing• Single layer• Room temp. substrate• Convective drying
Used to demonstrate new materials and for fundamental studies
Needed to demonstrate scalability of materials and MEA/cell designs, and industrial relevance
Scale Up
NREL | 6
Approach:Integrated Approach for Processes Scale-Up
Ink Characterization
RheologyDynamic Light Scattering
Zeta PotentialUSAXS
Electrode FabricationCoating Method
Drying Rate/TempSubstrate
Ex Situ Electrode CharacterizationElectron microscopy
X-ray tomography
In Situ Electrode Characterization
Fuel cell performanceImpedance spectroscopyTransport measurements
Ink FormulationCatalystIonomerSolvents
Dispersion method
Unique Aspects and Capabilities of this Project
NREL | 7
Approach: Project Schedule and Milestones
Qtr Date Milestone/Deliverable (as of 3/4/2019) Type Status
FY19Q3
6/2019 Determine influence of solvent formulation on ionomer adsorption on catalyst/support.
QPM MET
FY19Q4
9/2019 Evaluate ink formulations, drying conditions, and substrates to reduce crack formation in fuel cell and electrolysis catalyst layers coated using scalable methods.
QPM MET
FY20Q1
12/2019 Prepare catalyst layers with PtCo/HSAC (or other current advanced alloy catalyst) coated using scalable lab methods (Mayer rod and/or doctor blade) and/or continuous roll-to-roll methods (slot-die and/or gravure) for comparison to non-alloy catalysts using standard initial performance testing. Catalyst layers will also be provided to FC-PAD for advanced characterization and durability testing.
QPM MET
FY20Q2
3/2020 Characterize the influence of ionomer chemistry on catalyst-ionomer interactions in catalyst ink dispersions.
QPM MET
FY20Q3
6/2020 Characterize the influence of catalyst ink solids concentration on catalyst layer microstructure and performance for electrodes coated using scalable coating methods using at least two different catalyst-types (e.g. Pt/C, LTE, PGM-free) and three concentrations.
AM Delayed due to COVID
FY20Q4
9/2020 Explore the effect of additives (e.g. binder materials) and/or catalyst ink modifications (e.g. sample preparation methodology) on catalyst layer crack formation.
QPM Delayed due to COVID
This project is part of NREL’s Manufacturing R&D Project (TA001)
NREL | 8
, 50 cm2
Accomplishments and Progress:Utilizing Microscopy to Understand Influence of Coating Methods
2019 Fuel Cell Performance F/Pt = 1.8
F/Pt = 1.0
• Slot die coating results in higher performance than gravure• H2/N2 EIS showed performance difference due to differences in catalyst
layer resistance• TEM demonstrates coating method influences catalyst layer morphology• Slot die catalyst layer more porous likely due to lower shear rates during
coating• Gravure leads to less ionomer in catalyst layer, high shear rates may be
forcing ionomer into MPL• Lower ionomer in gravure coated CL consistent with EIS results.
NREL | 9
Accomplishments and Progress:Determined Influence of Solvent Formulation on Ionomer
Structure on Catalysts
Work done in collaboration with K.C. Neyerlin and FC-PAD
Dynamic Light Scattering Zeta Potential
Concentrated Ink Rheology
Liquid
Gel
• Addition of salt screens electrostatic repulsion due to SO3
- groups• No change in particle size for 62 wt% H2O indicates
strong steric stabilization – adsorbed ionomer is less compacted
• Concentrated ink with moderate water content has liquid-like rheology – well dispersed – consistent with DLS of dilute inks
• Catalyst ink structure consistent with catalyst layer transport properties and electrochemical performance
NREL | 10
Accomplishments and Progress:Development of Accurate Crack Measurement Algorithm
• Currently available methods/software (Otsu’s method, ImageJ, etc.) do not accurately detect cracks in catalyst layers due to heterogeneity of catalyst layer surface images
• NREL-developed algorithm reduces noise, thresholds, and filters image to accurately determine crack area
• Algorithm validated using control images with known crack areas and contrast
• Algorithm code is open-source (developed in Python)
Original Image Otsu’s Method1 NREL Method
Crack Area = 0.839%Crack Area = 56.5%
1) Otsu, IEEE Trans. Syst., Man, Cybern. 1979, 9(1), 62-66.
black pixels = “crack” black pixels = “crack”
Original Image Overlaid with NREL-Method Cracks
NREL | 11
Accomplishments and Progress:Determined that Decreased Agglomeration Leads to More Cracks
Gel-like (agglomerated)
Liquid like (dispersed)
• Studied loadings relevant to light-duty and heavy-duty fuel cells and low-temperature electrolysis• Increased agglomeration leads to decrease in cracks, due to decreased packing density, capillary
pressure during drying, and increased particle size• Results consistent with stress-limited drying physics:
high watercontent
ℎ𝑚𝑚𝑚𝑚𝑚𝑚 = 0.64𝐺𝐺𝐺𝐺𝜙𝜙𝑟𝑟𝑟𝑟𝑟𝑟𝑅𝑅3
2𝛾𝛾
1/2 2𝛾𝛾−𝑃𝑃𝑚𝑚𝑚𝑚𝑚𝑚 𝑅𝑅
3/2
Singh et al., Langmuir 2009, 25(8), 4284-4287
NREL | 12
Accomplishments and Progress:Completed ANOVA Analysis of Factors Influencing Cracking
ANOVA Analysis
F-Value Prob > F
I/C 1.33 0.29312
nPA 39.06 6.98 x 10-7
Rod Size 5.21 0.00699
I/C*nPA 3.47 0.03185
I/C*Rod Size 0.37 0.92123
nPA*Rod Size 1.98 0.11670
R2: 89%
• ANOVA statistical analysis shows 1-propanol content has highest F-value indicating strongest impact on cracking
• Inks that result in least cracked catalyst layers also lead to lower initial electrochemical performance, though a causal relationship has not been established
• Need to further explore the relationships between ink formulation, cracking and performance, especially as a function of loading/thickness
Crack % of Total Area
NREL | 13
, 50 cm2
Accomplishments and Progress:Demonstrated Performance Trends in R2R PtCo GDEs
Materials• Catalysts: Umicore PtCo/HSC• Ionomer: Nafion 1000 EW• Cathode Loading: 0.08 and
0.15 mgPt/cm2
• Membrane: 25 µm Nafion• GDL: Freudenberg H23C8
Coating Speed: 1 m/minInk: 7 wt% solidsWater/1-propanol Ratios: 75/25, 55/45, 25/75 w/w
• Collaboration with FC-PAD – R2R-coated materials provided to FC-PAD for advanced characterization and durability testing
• PtCo GDEs show same performance trend as Pt GDEs – increasing H2O increases performance
• Performance trends are the same in wide RH range (40-100 %RH)
0 0.5 1 1.5
i (A/cm2
)
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Ece
ll (V
)
25wt% H2
O
55wt% H2
O
75wt% H2
O
H2/air100% RH80 ˚C150 kPa
PtCo - 0.08 mgPt/cm2
Pt - 0.12 mgPt/cm2
Increasing H2O
Increasing H2O
NREL | 14
Accomplishments and Progress:Demonstrated Influence of Ionomer Equivalent Weight on Catalyst Inks
Studying effects of ionomer side chain length and spacing with three different ionomers
Ionomer Carbons in Side Chain
EW Side Chain Spacing
Nafion 5 1000 5.7
3M PFSA 4825 4.5
725 3.5
Aquivion 3830 5.5
720 4.4
Nafion
3M PFSA
Aquivion
• Decreasing EW leads to decrease in elastic modulus indicating an increase in dispersion of catalyst particles
• Low EW ionomers show liquid-like rheology (tan 𝛿𝛿 > 1)
• Rheology shows greatest dependence on sidechain length
• And less dependence on side chain spacing
https://www.sigmaaldrich.com/technical-documents/articles/materials-science/perfluorosulfonic-acid-membranes.html
Decreasing EW
Liquid
Gel
tan 𝛿𝛿 =𝐺𝐺𝐺𝐺𝐺𝐺
G’ – elastic modulusG” - viscous modulus
Rheology of Catalyst Inks with Ionomer of Different EW
NREL | 15
Accomplishments and Progress: Responses to Previous Year Reviewers’ Comments
• “…but the need is less clear. The original equipment manufacturers and electrode manufacturers have excellent electrode knowledge; it is not clear that the national laboratories should be trying to develop this knowledge.”– Some fuel cell OEMs may know how to make electrodes, but not
all do, and this information is not publicly available. Small-medium businesses do not have R2R process capabilities. Less understanding exists in the LTE space. The project also provides a knowledge and equipment platform supporting other national lab and university materials development.
• “It would be great to see the isotherm work on the catalyst inks as a function of ionomer loading and how that varies with the solvent system.”– We continue to work on this. This was part of our Q3 QPM for FY19.
• “The project team should add more industry interactions.”– This is being addressed through the AMO- and FCTO-funded R2R
Collaboration (TA007), with the collaboration of two U.S. fuel cell and two U.S. electrolyzer manufacturers
NREL | 16
Tech Transfer Activities:Leveraging Expertise in Projects with Industry
• Completed one Strategic Partnership Project and currently setting up another
• Lynntech (completed)• Optimized ink formulation for R2R coating using materials
supplied by Lynntech• Conducted roll-to-roll catalyst layer coatings• Provided coated materials to NREL colleagues for QC analysis
(TA001)• Coated materials provided to Lynntech
• Giner (upcoming)• Conduct rheological measurements of FC and LTE catalyst inks• Optimize coating and drying conditions to achieve R2R-coated
catalyst layers with target loading• Coated materials will be supplied to Giner
NREL | 17
Collaborations
Institution RoleNational Renewable Energy Laboratory - PrimeMichael Ulsh, Scott Mauger, Sunilkumar Khandavalli, Jason Pfeilsticker, Min Wang, Radhika Iyer, K.C. Neyerlin, Tim Van Cleve, Carlos Baez-Cotto
Ink formulation studies, electrode production and coating, rheology, MEA performance testing, advanced diagnostics
Argonne National LaboratoryDebbie Myers, Jae Hyung Park, Nancy Kariuki
Small angle x-ray scattering of catalyst inks – critical for understanding rheology measurements and catalyst ink microstructure
Colorado School of MinesSvitlana Pylypenko, Samantha Medina
Electron microscopy of catalyst materials and electrodes
3MAndrew Haug, Mike Yandrasits
Ionomer powders and dispersions
NREL | 18
Remaining Challenges and Barriers
• Improve understanding of effects of ionomer chemistry and processing on interactions with catalysts
• Extend cracking studies to LTE materials• Cracking tends to be more prevalent in ink
formulations that result in higher electrochemical performance. Need to understand if (and how) these are related.
• Develop understanding of how catalyst ink concentration influences processability, performance, and cracking behavior
NREL | 19
Proposed Future Work
• Extend cracking studies to IrO2 catalysts• Explore strategies (additives, processing) to mitigate
cracking in catalyst layers• Perform in situ testing studies to better understand
relationships between cracking and electrochemical performance
• Continue studies on influence of ionomer chemistry on catalyst ionomer interactions
• Study influence of catalyst concentration on processability, microstructure, and electrochemical performance of fuel cell and LTE catalyst layers
Any proposed future work is subject to change based on funding levels
NREL | 20
Summary
Objective: Study material-process-performance relationships for R2R PEMFC/EC cell materials to understand relationships between process science and material properties and performanceRelevance: Addressing MYRD&D milestones. This project is enabling for other DOE-funded researchApproach: Understand impacts of ink formulation, ionomer EW, and catalyst type on ink microstructure, electrochemistry, cracking behavior, and catalyst – ionomer interactionsAccomplishments: • Determined that slot-die coating results in more porous catalyst
layer structure than gravure coating• Developed open-source script for analyzing catalyst layer cracking• Determined that higher nPA content in catalyst ink lowers crack
percentage• Initiated work effects of ionomer EW in inks – lower EW leads to
better dispersion of Pt/C particles
Technical Back-Up Slides
NREL | 22
, 50 cm2
Accomplishments and Progress (2019):Characterized Influence of Coating Method on Performance
Method ECSA (m2/gPt)
im0.9V
(mA/mgPt)
Slot die 63.1 ± 2.7 395 ± 21
Gravure 64.5 ± 2.3 395 ± 22
Fuel Cell Performance
• EIS used to understand differences in proton conductivity
• Model fitting and analysis shows gravure results in higher effective catalyst layer resistance
• Slot die coating results in higher performance than gravure, regardless of ink formulation
• Coating method does not impact kinetics or site accessibility
• Polarization curves suggest performance differences are due to differences in Ohmic losses between MEAs
• May also be differences in transport – needs further exploration
H2/N2 Electrochemical Impedance Spectroscopy
Δ𝐸𝐸 = 23.6 𝑚𝑚𝑚𝑚
∆𝑅𝑅𝐶𝐶𝐶𝐶 = 33.1 𝑚𝑚Ω � 𝑐𝑐𝑚𝑚2
∆𝐸𝐸 = 𝑖𝑖∆𝑅𝑅𝐶𝐶𝐶𝐶∆𝐸𝐸 0.8 𝐴𝐴
𝑟𝑟𝑚𝑚2 = 26.5 𝑚𝑚𝑚𝑚
Model Fit